The definition of RPDIV in impulsive testing and the effect of the electrical conditioning on the variability of the results

This paper investigates the effect of preconditioning on the variability of the results when measuring the Repetitive Partial Discharges Inception Voltage (RPDIV) of electrical insulation systems subjected to impulsive voltage stresses. The reference International Standard considered in the following (IEC 60034-18-41 [1]) defines the RPDIV as the minimum voltage at which at least five Partial Discharges (PD) occurs in ten voltage impulses applied. Many testing devices have been designed according to such definition: they inject ten voltage impulses and, if five of them trigger PD events, the voltage is recorded as the RPDIV. Otherwise, the voltage is increased until the RPDIV condition is met. On the other hand, other systems focus their analysis on the time behavior of the PD probability in the course of the tests. They do not simply record the RPDIV at the ratio where the number of PDs is greater than 0.5 times the voltage surges, but when it is persistently larger than such threshold. It has been noted that the PD probability (i.e. the number of PD per number of voltage pulses) shows a non-monotone behavior during the RPDIV measurements and consistent variations above and below the 50% threshold occur. Therefore, results coming from the types of equipment having the two different approaches are often noticeably different and, as a consequence, the RPDIV results using different measurement systems might vary depending on how this parameter is calculated. This paper proposes the electrical conditioning as a possible procedure to reduce this variation, and thus to have results independent from the measuring apparatus.